Validating an Objective Roadside Tool to Assess Driver Fatigue

Principal Investigator:

Nichole Morris, Research Associate, Mechanical Engineering


Project Summary:

Fatigue not only impairs reaction time, attention, and risk assessment, but leads to increased micro-sleeping, falling asleep at the wheel, and lane drifting, all of which negatively affect safe driving. NHTSA conservatively estimates 37,000 injuries and 900 deaths caused by drowsy driving each year in the U.S., making it a leading preventable cause of road deaths. In 2015, 482 crashes were attributed to drowsy driving in Minnesota; however, these rates may be underestimated given that law enforcement lack an objective tool to determine fatigue-involved crashes. Moreover, AAA found in a recent study of dashcam recordings that nearly 10% of crashes are attributed to drowsy driving. Commercial truck drivers, who often work irregular, long shifts under monotonous conditions, are overrepresented in fatal sleep-related crashes. Similar work schedules are also common among snowplow drivers who must respond to demanding snow emergencies, not bound by HOS regulations, which can further exacerbate fatigue. These findings highlight the need for an objective roadside tool to detect driver fatigue. Notably, law enforcement officers (e.g., MN State Patrol) currently lack an objective tool to assess fatigue, among other types of impaired driving. The project aims to validate critical flicker frequency (CFF), a reliable and valid psychophysical technique that measures an individual's level of alertness, as an objective roadside tool to assess fatigue. The goals of this project are: 1. Collect a large sampling of CFF responses to compare to self-reported sleep schedule 2. Develop a sensitive driving measurement test for driver fatigue that can be applied to other types of impairment 3. Validate the accuracy of the measurement tool to assess impairment among drivers who are experiencing acute driver fatigue (i.e., 34 hours of prolonged wakefulness).


Project Details: